About the Role
You will curate fleet data, build enrichment and labeling pipelines, train and fine-tune large-scale models, and design offline and closed-loop evaluation. You will develop methods for embodied vision-language-action models, world models, policy learning, reinforcement learning, and reward modeling.
Requirements
•Master's degree with around 6 or more years of relevant experience, or PhD with 2 or more years
•Machine learning and software fundamentals
•Experience taking machine learning research into production systems at scale
•Experience in data curation, foundation-model training, large-scale data wrangling, or foundation-model evaluation
•Experience with large-scale data or large neural networks
•Python proficiency
•Experience with PyTorch or a similar deep-learning framework
Responsibilities
•Mine fleet data for rare, long-tail, and safety-critical events
•Develop repeatable training-data curation across cities, sensor rigs, and embodiments
•Build automated and semi-automated enrichment, labeling, and data-quality pipelines
•Build and fine-tune large-scale pretrained models
•Develop embodied vision-language-action models for driving
•Design offline and closed-loop evaluation metrics and benchmarks
•Use world-model-based evaluation for counterfactual scenarios
•Contribute to policy learning, reinforcement learning, and reward modeling
Benefits
•Hybrid working
•Competitive equity package
